Diabetes and Dyslipidemia: Do Recent Trials and Guidelines Improve Our Treatment Strategies?

At an industry-sponsored symposium held in conjunction with the American Association of Clinical Endocrinologists’ 2003 Annual Meeting and Clinical Congress, two leaders in the diagnosis and treatment of diabetes and associated dyslipidemias presented the latest information on the pathophysiology and epidemiology of type 2 diabetes as it relates coronary artery disease (CAD) and therapeutic goals in diabetic dyslipidemia. Topics included the pathogenesis of increased atherosclerosis in the diabetic patient, metabolic factors that contribute to early onset of complications and clinical trial evidence supporting the effectiveness and safety of lipid modification therapies.

This program was supported by an unrestricted educational grant from Merck and Co., Inc.

The Pathophysiology and Epidemiology of Type 2 Diabetes in Relation to CAD

Recent studies have proven that type 2 diabetes (T2DM) has such a profound effect on coronary artery disease (CAD) that it is now seen as an independent risk factor for heart disease. The reasons for this, and what can be done to lessen the incidence, are just now becoming clear.

Andre Krolewski and colleagues compared the cardiovascular (CV) mortality in patients with T2DM at the Joslin Clinic and a nondiabetic cohort from the Framingham Heart Study. There was a twofold increase seen in men and a fivefold increase of cardiovascular death in women (Krolewski AS, et al. Am J Med. 1991;90 (suppl 2A):56S).

Similar findings were seen in other studies. In a Finnish population, patients with diabetes and no prior history of myocardial infarction (MI) had the same risk of a heart attack as the nondiabetic who had a previous MI (Haffner S, et al. NEJM. 1998;339:229).

Results from the multi-national Organization to Assess Strategies for Ischemic Syndromes (OASIS) study suggest that these concerns are not limited to Finland. In this cohort, those with diabetes and no history of CAD had the same 70% increase in CV mortality as nondiabetics with a previous MI (Malmberg K, et al. Circulation. 2000; 102:1014).

“These are indications of how serious this problem is,” said Om P. Ganda, MD, Associate Clinical Professor of Medicine, Harvard Medical School in Boston. “This data has lead to a reawakening among clinicians and new guidelines that emphasize diabetes is a serious risk equivalent for cardiovascular events.

“We now know that there are multiple pathways through which hyperglycemia could be responsible for increasing macrovascular complications,” continued Dr. Ganda. “Patients with diabetes have increased atherogenesis not only from increased blood glucose and lipid disorders, but perhaps also from insulin resistance.”

Other studies indicate that frank diabetes may not be required to see changes in CAD indicators. Work by Haffner and his associates showed that there was an increase in CV risk factors present prior to development of T2DM. A group of healthy subjects was followed for 8 years and those who became diabetic over that time already had higher triglycerides, lower LDL, higher blood pressure, and increased fasting blood glucose (FBG) levels at baseline. Their fasting insulin levels were twice as high as those who did not eventually develop diabetes (Haffner, SM, et al. JAMA. 1990;263: 2893).

“The processes of cardiovascular disease (CVD) are already in play for many years before a clinical diagnosis of T2DM,” said Dr. Ganda. “Similar data from the Women’s Health Initiative helps to confirm this finding.”

Insulin resistance syndrome (IRS) may also be a player in CVD. The Veterans Affairs High-density Lipoprotein Intervention Trial (VA-HIT) looked at fasting insulin quartiles in subjects without diabetes. Those in the higher quartile of insulin had the greatest benefit when given gemfibrozil (Rubins HB, et al. Arch Intern Med. 2002;162: 2597).

Lipid abnormalities in T2DM include low LDL cholesterol and high HDL. In addition, there is an increase in remnant particles, small and dense LDL, cholesterol-enriched VLDL, triglyceride enriched HDL, and glycosylation of apoproteins and phospholipids. These qualitative defects, according to Dr. Ganda, may be at least as important as the quantitative defects when explaining the increased risk of atherosclerosis.

In recent data from the Framingham Offspring study, the most important lipid problem in patients with T2DM was low HDL cholesterol. Using the old cutoff of 35 mg/dl, twice as many men with diabetes have low HDL when compared to those without. High triglycerides (> 250 mg/dl) were seen at about the same level. The differences were magnified in women with diabetes where there were four- and six-fold increases, respectively in the two measures (Siegel, et al. Metabolism.1996;45:1267).

Results from a study by Dr. Tim Garvey and colleagues, suggested that IR has an effect on lipoprotein particle size. As the patients went from insulin sensitivity to insulin resistance to diabetes, the VLDL becomes larger and LDL becomes small, both considered to be atherogenic. HDL is almost as small in those with IR as seen in those with T2DM (Garvey WT. Diabetes. 2003;52: 453).

“The problems with these compositional defects in lipoproteins begins very early on as insulin resistance develops,” said Dr. Ganda.

Currently, there is controversy over the independent contributions of triglycerides in patients with diabetes. Austin and others have shown that for every 90 mg/% increase in triglycerides, there is an increase in HDL-adjusted risk of coronary event of 14% in men and 35% in women. Dr. Ganda noted that there are currently no clinical trials published that have proven triglyceridemia per se is an important risk factor (Hokanson JE, et al. J Cardiovascular Risk. 1993;3: 213).

“When it comes to HDL, I think we again have a lot to learn about ways to increase HDL and whether it is an independent predictor of CV events,” said Dr. Ganda. “On the other side, we also might ask ourselves about treatment goals for LDL cholesterol.”

“Adult Treatment Panel III (ATP-III) guidelines suggest a goal of 100 mg/dl or less for LDL cholesterol,” said Dr. Ganda. “The important point ATP-III makes is that simultaneous initiation of nonpharmacological measures along with LDL lowering drugs is needed in patients with LDL cholesterol levels above 130. If the LDL target is met and triglycerides are > 200, then non-HDL cholesterol becomes the secondary target.”

Non-HDL assessment allows the clinician to look at all the Apo B containing particles such as LDL, remnant particles, VLDL and Lipoprotein(a) (Lp(a)). There are a number of studies looking at the question of non-HDL cholesterol vs. LDL cholesterol as a risk predictor.

The Atorvastatin Comparative Cholesterol Efficacy and Safety Study (ACCESS) looked at the correlation between non-HDL and LDL cholesterol with Apo B. The correlations were better in this study with non-HDL cholesterol in relation to Apo B as compared to LDL/Apo B (Ballantyne CM, et al. Am J Cardiol. 2001;88:265).

Another example is the long-term follow-up of the Lipid Research Clinics Trial (LRC). As the LDL cholesterol levels increased, all the risk lines crossed zero meaning they were not significant. Non-LDL cholesterol is progressively more likely to give the better risk ratio (Cui Y, et al. Arch Int Med.2001;161; 1413).
However, based on recent evidence from the large AMORIS database and other studies, Allan Sniderman contends that non-HDL cholesterol is not as good a predictor of atherogenic risk as the direct measurement of Apo B. This may be particularly relevant in T2DM where alterations in the LDL and IDL composition result in underestimation of true Apo B particle concentrations (Sniderman AD, et al. Diabetes Care. 2002;25: 579).

It is important to appreciate that while a low HDL is considered a risk factor, there are no ATP-III goals set for HDL cholesterol,” Dr. Ganda. “There is a goal set for HDL by the American Diabetes Association (ADA) and maybe it is time to reconcile the two.”

The ADA Position Statement suggests that treatment, in either gender, should be aimed at getting HDL above 40 mg/dl if possible. Some think that 45 or even 50 may be better. How to achieve that remains an open question.

Another possible method to refine risk assessments would be to look at various ratios of LDL, HDL, and triglycerides. Dr. Michael Gaziano and coworkers looked at 340 cases of MI and matched them with 340 others as controls. Their data showed that after adjusting for all other risk factors, triglycerides alone accounted for a two-fold increase in risk. However, there were much greater increases when they assessed the triglyceride/HDL ratio (Gaziano JM, et al. Circulation 1997;96: 2520).

Metabolic syndrome or insulin resistance can also have an effect on cardiac health. Since up to 25% of adults in the United States (perhaps half of those over the age of 50) have metabolic syndrome/insulin resistance, finding and treating these individuals is important.

Alexander and colleagues looked at the prevalence of CHD in relation to metabolic syndrome. Those with diabetes without metabolic syndrome had a prevalence of 7.5%. Subjects with the syndrome had the highest prevalence (14-19%) of CHD independent of diabetes status (Alexander CM, et al. Diabetes. 2003;52:1210) (Figure 1).

The importance of diagnosing this syndrome is illustrated by results from Kupio Heart Study. The researchers prospectively enrolled 1,200 Finnish men between 42 and 69 and then followed them for 12 years. They measured CHD, total cardiovascular, and all cause mortality (Lakka H-M, et al. JAMA. 2002; 288:2709).

“In each case, there was a two- to three-fold greater risk of mortality based simply on metabolic syndrome status,” said Dr. Ganda. “We have suspected that these people have an increased risk and we are now seeing the studies that back it up.”

The Insulin Resistance and Athero-sclerosis Study may be giving physicians some clues on why the two are linked. Results indicate that inflammatory markers of endothelial dysfunction were good predictors of T2DM and consequently CVD. Fibrinogen, C-Reactive Protein (CRP), and plasminogen activator inhibitor-1 (PAI-1) were progressively associated with higher risks of diabetes (Festa A, et al. Diabetes. 2002;51:1131). Similar results were seen in the Women’s Health Initiative, where both Il-6 and CRP were predictors for diabetes, and VA-HIT where an association with insulin resistance transcends into inflammatory response.

The class of medications known as fibrates may be useful in treating these concerns. A recent study by Staels, et al, has shown that PPAR-alpha agonists such as fenofibrate lowered not only triglycerides, but also fibrinogen, CRP, and Il-6 (Barbier O, et al. Arterioscler Thromb Vasc Biol. 2002;22:717).

“There is a lot of debate now on when to measure CRP and what to do about it,” said Dr. Ganda. “Widespread screening probably isn’t appropriate and most patients with diabetes already have high levels. It is only in those with metabolic syndrome and intermediate risk according to the Framingham
equations that should be considered for CRP testing.”

Pharmacologic Intervention and Clinical Trial Evidence: Therapeutic Goals in Diabetic Dyslipidemia

One of the major questions facing clinicians is the relationship between cardiovascular events, various indicators of diabetes and the lipid dysfunctions that accompany it.

Turner and colleagues looked at the results of the United Kingdom Prospec-tive Diabetes Study (UKPDS). Then they entered a series of variables into a multivariate model, LDL was first followed by HDL, hemoglobin A1C, systolic blood pressure, and smoking (Turner RC, et al. BMJ. 1998;316:823). “This was a surprise to many of us at the American Diabetes Association,” said Steven M. Haffner, MD, Professor, Department of Medicine, University of Texas Health Science Center at San Antonio. “The usual mantra concerning treatment of diabetic dyslipidemias had been to treat high triglycerides and low HDLs because the LDLs were frankly not significant. This helped to reestablish the order of priorities.”

In the late 90s, much data on secondary prevention was being generated. What was lacking was data in the primary prevention area. The Heart Protection Study (HPS) broke new ground by establishing that statin therapy worked as a primary prevention tool. They enrolled 20,500 subjects and of these 6,000 subjects were type 2 diabetics. The two interventions were 40 mg simvastatin or placebo. There was also an antioxidant arm that failed to show an effect.

In the first year, the results were not statistically significant, but over half of the full effect was seen in this time period meaning that the changes do begin to occur early. It reached significance by the end of the second year and the treatment effect does not change from year-to-year afterwards. (Collins, R, et al. Lancet. 2002;360:7).

“There were so many subjects at such a high risk that the Kaplan-Meier curves were deceptively smooth,” said Dr. Haffner. “There was a separation, but it occurred continuously.”

Most of the focus was on vascular events such as MI, strokes, and revascularizations. There was a 25% reduction in any major vascular event. Nonfatal incidence and coronary death were similar to coronary events showing no evidence of heterogeneity.

Stroke rates were also reduced by about the same amount as major coronary events. This surprised the researchers since cholesterol is a weaker predictor of stroke than it is of MI.

“This is partially because we had no confidence that statins would reduce stroke,” said Dr. Haffner. “But clearly it did in this case and with no adverse effect on hemorrhagic stroke.”

The HPS also looked at the effect of simvastatin on first major vascular event according to baseline LDL cholesterol levels. They found significant reductions in events even in those who started below goal. That has lead some to suggest that people should be placed on statins even if they start with an LDL below 100 mg/dl.

“The conventional view of trials breaks them into either primary or secondary prevention trials,” said Dr. Haffner. “The newer generation of trials such as ALLHAT and the HPS, have both included in their design.”

Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT) had 10,000 subjects enrolled in the lipid intervention arm who were randomized to pravastatin 40 vs. placebo. In this particular trial there was only about a 9% differential in cholesterol with a similar reduction in coronary heart disease.

“This looks like a bad trial to many people, but I think it shows that in order to get a reduction in CHD, you actually have to have significant reductions in events,” said Dr. Haffner. “There is nothing magical about just being on a statin.”

The Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT) was principally a hypertension study of 10,000 subjects randomized to 10 mg atorvastatin vs. placebo. Of these, 2,500 were diabetic. Overall, all cause mortality was reduced by 13%. The reduction in diabetic subjects was only half as large as that seen in nondiabetic and was not statistically significant. (Sever PS, et al. Lancet.2003;361:1149).

“In the discussion, they mentioned that 14% of those in the placebo arm crossed over to active in the diabetics compared to only 8% overall,” said Dr. Haffner. “One thing you might expect is that there may have been a smaller differential in LDL cholesterol among the diabetic subjects which might explain some of the lessened effect.”

They also looked at those with and without metabolic syndrome. The differences did not reach significance in those with the syndrome, but it did in subjects without. Dr. Haffner attributed this to problems with powering.

The STENO II study looked at very aggressive treatment with Angiotensin Converting Enzyme Inhibitors (ACE-I) and Angiotensin Receptor Blockers (ARB). While the people in the intensive treatment group were more likely to reach their goal of A1C levels less than 6.5%, both groups did poorly. Looking at composite endpoints, there was a 53% reduction in risk in the intensive group (Gaede P, et al. NEJM.2003;348:383).

“Many wonder how we get more than the 25% or 30% reduction seen in the best statin or ACE-I trials,” said Dr. Haffner. “I think this study shows that multi-risk factor interventions work.”